The present disclosure generally relates to ceramic collars for active brazing to metal rings in sodium-based thermal batteries, and methods of active brazing ceramic collars to metal rings in sodium-based thermal batteries.
High-temperature rechargeable batteries, such as sodium-based thermal batteries like sodium metal halide or sodium sulfur cells, typically have a number of components that need to be sealed for the cell to work. Sodium metal halide (NaMx) batteries, for instance, may include cells including a sodium metal anode and a metal halide (NiCl2 for example) cathode. A beta″-alumina solid electrolyte (BASE) separator can be used to separate the anode and cathode. The solid electrolyte may allow the transport of sodium ions between anode and cathode. A secondary electrolyte (NaAlCl4) can also used in the cathode mixture. The cathode mixture typically consists of nickel and sodium chloride along with other additives. The cathode mixture is contained inside the BASE tube, which is closed on one end.
In the present typical design of NaMx cells, the open end of the beta″-alumina ceramic tube is joined to an alpha-alumina collar using a glass seal. Spinel, zirconia, yttria, or other ceramic insulators, or combinations thereof, may also be used as a collar material in NaMx cells. The alpha-alumina collar isolates electrically the anode from the cathode. In order to enable the coupling (e.g., welding) of this ceramic subassembly to the metallic current collectors anode and cathode), two metallic rings (typically Ni) are coupled or otherwise bonded or sealed (e.g., hermetically sealed) to the alpha-alumina collar prior to the sealing glass operation. The inside metal ring is typically welded to the cathode current collector and the outside metal ring is typically welded to the anode current collector (e.g., the battery case). The integrity of these two metal-ceramic joints is critical for the reliability and safety of the NaMx cell. This conventional, sealing arrangement of an alumina collar structure in these types of cells, relative to one or more metal rings, is depicted in FIGS. 1, 2A, and 2B, which are further described in detail below. FIG. 1 is a perspective sectional view of a portion of an exemplary NaMx cell, including a prior art, exemplary alpha-alumina collar, sealed to nickel (Ni) outer and inner rings. FIGS. 2A-B are perspective views of the sealed prior art alpha-alumina collar and Ni rings of the cell of FIG. 1.
The coupling of the metal (e.g., Ni) rings and ceramic (e.g., alpha-alumina) collar is typically achieved with two main process steps: (1) metallization of the alpha-alumina collar; and (2) thermal compression bonding (TCB) of both Ni rings to the metalized ceramic collar. Generally speaking, the first process step of metalizing the alpha-alumina collar provides a bond (e.g., a glass bond) between a pure Mo metallization layer and the alumina collar, and the second process step of TCB provides a diffusion bond between the Mo in the metallization layer and the Ni of the inner and outer rings.
As mentioned above, to be able to join a Ni ring via a TCB to an alpha-alumina collar in a NaMx cell, it is necessary to initially metalize the alumina. Without the metallization, it is difficult to create a metallurgical bond during the TCB process between the Ni ring and the alpha-alumina collar. Metalizing of alumina has been practiced since the late 1940's, with the Mo—Mn process being the most studied and the most widely commercialized metallization process for alpha-alumina. In the process, the paste material is applied to alumina typically via screen printing, and heated treated (e.g., about 1500 degrees C. to about 1600 degrees C.) with wet hydrogen to bond the Mo to the alumina. During the heating process glass flows from the debased alumina into the Mo layer, and the wet hydrogen may promote the wicking and wetting of the glassy phase in the alumina into the Mo layer. However, in a NaMx cell Mn is incompatible with the chemistry used in the cell and is highly susceptible to corrosion. It is therefore necessary to use a metallization process that uses only a 100% Mo paste. Unfortunately, using 100% Mo makes the metallization process more difficult and narrows the process window by significantly restricting the operating ranges of common processing variables, temperature, dew point, and glass composition. Further TCB process is a batch-process and requires large investments to produce large number of parts. As a result, the TCB process is also time consuming, not-scalable and expensive.
Once formed, however, the Mo metallization layer provides a metal surface for the bonding of the Ni rings to the alpha collar. As mentioned above, the Mo layer is a composite comprised of two interlaced phases—Mo and glass. The subsequent thermal compression bonding (TCB) step is the formation of a metallurgical bond between the Ni ring and Mo metalized layer on the alpha-alumina collar. Specifically, the bond is created by heating the Ni rings and metalized alumina collar while they are in contact and relatively high pressure is applied to the joints therebetween. To create a sufficient bond, the Ni rings and metalized alumina collar must be subjected to relatively high temperatures (e.g., at least about 950 degrees C.), for relatively long periods of time (e.g., at least about 45 minutes) and while subjected to a significant load (e.g., at least about 750 kg force). Further, each Ni ring and alumina collar subassembly must be individually arranged or processed such that the Ni rings are properly located on the alumina collar and the compressive load is applied to the joint between the Ni rings and alumina collar. Thereby, the TCB process is also time consuming and expensive.
Although the metallization and TCB process is difficult, time consuming and expensive, it is the typical process to bond Ni rings and alpha-alumina collars in NaMx cells due to the relatively high bond strength achieved thereby. In fact, the main advantage or CTQ (Critical to Customer) parameter of the metallization and TCB sub-assembly is the metal-to-ceramic bond strength achieved between the Ni rings and the alpha-alumina collar, along with hermeticity of the bond. Typically, the metal-to-ceramic bond between the Ni rings and the alpha-alumina collar are tested by a peel test which subjects the metal-ceramic bond to a tensile stress until failure while the load-to-failure variable or metric is measured. While the tensile strength of the metal-to-ceramic bonds are important (such as to sufficiently withstand internal pressures present during the operation of NaMx cell batteries), it is noted that the tensile strength of the bonds is used as an overall strength indication of bonds (i.e., ability to withstand tensile and other forces present during the operation and life of NaMx cell batteries).
The strength of the TCB bond on both the inner and outer rings is controlled by a wide range of variables inherent to the components of the TCB subassembly to ensure sufficient bond strength is achieved. The microstructure of the alpha-alumina collar and the Mo metallization, along with the TCB process heavily influence the final strength of the metal-to-ceramic bond. With upwards of forty different processing steps needed to manufacture the TCB subassembly, it is necessary to develop a quality control plan for all components of the subassembly. Again, the process to achieve the TCB subassembly (metalized alumina collar and TCB collar and Ni rings) is thereby difficult to achieve, non scalable, expensive and time consuming. As a result, to advantageously avoid the difficulties, expense and time associated with the metallization and TCB process typically associated with the manufacturing of NaMx cells, alternate joining technologies for the Ni rings and alpha-alumina collar that achieve sufficient bond strength are necessary.
One potential alternative joining technology or process for sealing or bonding the Ni rings and the alpha-alumina collar in NaMx cells is active brazing. For example, active brazing the Ni rings and the alpha-alumina collar may be capable of reducing NaMx battery costs by at least two dollars per cell as compared to current metallization and TCB technologies or processes.
Active brazing metal-ceramic joints is a procedure in which one of the components from a braze alloy reacts with the ceramic and forms an interfacial bond. Conventionally, brazing is done through metallization in combination with a braze alloy. One requirement of a braze alloy for use in high temperature rechargeable batteries, such as NaMx batteries, is high corrosion resistance towards sodium and halide. Active brazing has been known to join ceramic to metal, but there are not many commercially available active braze alloys (ABAs) suitable for use in NaMx cells. Specifically, high temperature ABAs (e.g., about 900-1200 degrees C.) and ABAs resistant to corrosion from sodium and halide, as required in NaMx cells, are in short order. Further, due to the high temperatures present during active brazing in NaMx cells and the significant mismatch of the coefficient of thermal expansions of the Ni rings and the alpha-alumina collar, the typical bond strength (e.g., tensile strength) achieved with prior art Ni rings and alpha-alumina collar designs by active brazing with suitable ABAs is commonly insufficient (i.e., sufficient bond strength is difficult to achieve by active brazing prior art Ni rings and alpha-alumina collar designs).
There continues to be a growing need in the art for high performance metal halide batteries with lower fabrication costs. Thus, Ni rings and alpha-alumina collar designs capable of being sealed or bonded via active brazing that exhibit sufficient bond strength (i.e., are capable of achieving typical NaMx batter performance) is desirable. For example, Ni rings and alpha-alumina collar designs effective in producing relatively minor residual stresses via active brazing on the alumina collar due to the thermal expansion mismatch between the ceramic collar and the metal (e.g., Ni) rings are advantageous. As another example, Ni rings and alpha-alumina collar designs effective in increasing bond strength (e.g., gaining a mechanical or structural advantage) between ceramic collar and the Ni rings at least in the tensile direction are advantageous. Such improved Ni rings and alpha-alumina collar designs for active brazing should provide for bond strengths at least comparable to bond strengths achieved with conventional metallization and TCB processes, reduce manufacturing costs compared with conventional metallization and TCB processes and/or reduce manufacturing times compared with conventional metallization and TCB processes.